The only surviving witness to the accident was also the least experienced mariner on board the DALEWOODPROVIDER, and he was asleep for some time before the accident. Consequently, he either did not notice, or did not remember, the circumstances leading to the capsizing. Based on the scanty information he provided regarding the immersion of the after end of the weather deck, it may be deduced that there was an ingress of sea water into some of the after underdeck spaces and the engine-room. The ingress could have happened gradually while the skipper was asleep, without either of the two deck-hands noticing it, or alternatively, suddenly, shortly before the surviving deck-hand was awakened. It is unknown if the skipper or the deck-hand who conned the vessel inspected the after compartments or the engine-room at any time during the trip in the rough weather conditions, but it is considered doubtful that they did. Reportedly, the skipper and his crew were awake for more than 20 hours prior to departure. The preparations for departure, the problems with the autopilot and the departure itself added to their fatigue, causing the skipper to go to bed immediately after the vessel left Sooke. It is conceivable that he did not ensure that all weather deck doors, hatches and manhole openings were completely secured. Neither deck-hand had enough sea-going experience and safety awareness in this respect to do so on his own. The inexperienced and tired wheelsman, while conning the vessel for several hours, would probably not have noticed any difference in the vessel's behaviour if water had gradually entered the hull. Alternatively, as the sole person at the con lacked navigational experience, the vessel could have deviated from her intended course, approached the unmarked and rugged coast, and struck the rocky bottom, causing damage to the hull and a sudden ingress of water. When the skipper first entered the wheel-house at about 0800, he was busy making a phone call, which indicates that the electrical power was available at that time. Because he did not survive, it could not be determined if he had noticed a gradual accumulation of water in the engine-room and after spaces, or if a sudden ingress occurred while he was in the wheel-house. The vessel appeared to be sinking at about 0900 when the suddenly awoken survivor saw that the after end of the weather deck was under water. The fact that the main engine was still running while there was no electrical power suggests that the batteries were flooded before the engine air intake was submerged, a sequence which is consistent with their relative locations. At the time of the capsizing, the vessel was not engaged in any fishery, and significant items of fishing equipment had been removed. Furthermore, unreported hull-form modifications to the bow and stern before and after the vessel's previous capsizing in 1989, together with the non-implementation of proposed remedial actions after that casualty, made the original as-built and CCG-approved Trim and Stability Booklet, and later revised stability data, redundant or inapplicable. The lack of records or other information regarding the taking-on or location of any compensating or trimming water ballast while the deck cargo of split cedar wood was unloaded precluded any post-casualty calculation of reliable, intact stability characteristics for the vessel on departure and also just before the capsizing. Regulatory requirements applicable to small fishing vessels of this class and to cargo ships of this size address only intact stability with no criteria for damaged conditions. Consequently, these vessels are highly vulnerable in the event of undetected flooding. The survivor's descriptions of the vessel's trim just before the capsizing, and of the actual capsizing, are consistent with the DALEWOODPROVIDER having taken on water in the after underdeck compartments and engine-room. The location, weight and free surface effect of this accumulated water were such that the after freeboard, reserve buoyancy and transverse stability were markedly reduced, and the vessel's remaining righting ability was overcome in the prevailing rough weather conditions. The exact position of the capsizing could not be established. However, it may be deduced that if the vessel was off Nitinat at 0800 when the skipper talked to his wife, and was proceeding in rough weather in a north-westerly direction, then, between 0900 and 0930, the vessel would have been somewhere off Pachena Point light (48o43.3'N, 125o05.8'W).Analysis The only surviving witness to the accident was also the least experienced mariner on board the DALEWOODPROVIDER, and he was asleep for some time before the accident. Consequently, he either did not notice, or did not remember, the circumstances leading to the capsizing. Based on the scanty information he provided regarding the immersion of the after end of the weather deck, it may be deduced that there was an ingress of sea water into some of the after underdeck spaces and the engine-room. The ingress could have happened gradually while the skipper was asleep, without either of the two deck-hands noticing it, or alternatively, suddenly, shortly before the surviving deck-hand was awakened. It is unknown if the skipper or the deck-hand who conned the vessel inspected the after compartments or the engine-room at any time during the trip in the rough weather conditions, but it is considered doubtful that they did. Reportedly, the skipper and his crew were awake for more than 20 hours prior to departure. The preparations for departure, the problems with the autopilot and the departure itself added to their fatigue, causing the skipper to go to bed immediately after the vessel left Sooke. It is conceivable that he did not ensure that all weather deck doors, hatches and manhole openings were completely secured. Neither deck-hand had enough sea-going experience and safety awareness in this respect to do so on his own. The inexperienced and tired wheelsman, while conning the vessel for several hours, would probably not have noticed any difference in the vessel's behaviour if water had gradually entered the hull. Alternatively, as the sole person at the con lacked navigational experience, the vessel could have deviated from her intended course, approached the unmarked and rugged coast, and struck the rocky bottom, causing damage to the hull and a sudden ingress of water. When the skipper first entered the wheel-house at about 0800, he was busy making a phone call, which indicates that the electrical power was available at that time. Because he did not survive, it could not be determined if he had noticed a gradual accumulation of water in the engine-room and after spaces, or if a sudden ingress occurred while he was in the wheel-house. The vessel appeared to be sinking at about 0900 when the suddenly awoken survivor saw that the after end of the weather deck was under water. The fact that the main engine was still running while there was no electrical power suggests that the batteries were flooded before the engine air intake was submerged, a sequence which is consistent with their relative locations. At the time of the capsizing, the vessel was not engaged in any fishery, and significant items of fishing equipment had been removed. Furthermore, unreported hull-form modifications to the bow and stern before and after the vessel's previous capsizing in 1989, together with the non-implementation of proposed remedial actions after that casualty, made the original as-built and CCG-approved Trim and Stability Booklet, and later revised stability data, redundant or inapplicable. The lack of records or other information regarding the taking-on or location of any compensating or trimming water ballast while the deck cargo of split cedar wood was unloaded precluded any post-casualty calculation of reliable, intact stability characteristics for the vessel on departure and also just before the capsizing. Regulatory requirements applicable to small fishing vessels of this class and to cargo ships of this size address only intact stability with no criteria for damaged conditions. Consequently, these vessels are highly vulnerable in the event of undetected flooding. The survivor's descriptions of the vessel's trim just before the capsizing, and of the actual capsizing, are consistent with the DALEWOODPROVIDER having taken on water in the after underdeck compartments and engine-room. The location, weight and free surface effect of this accumulated water were such that the after freeboard, reserve buoyancy and transverse stability were markedly reduced, and the vessel's remaining righting ability was overcome in the prevailing rough weather conditions. The exact position of the capsizing could not be established. However, it may be deduced that if the vessel was off Nitinat at 0800 when the skipper talked to his wife, and was proceeding in rough weather in a north-westerly direction, then, between 0900 and 0930, the vessel would have been somewhere off Pachena Point light (48o43.3'N, 125o05.8'W). Sea water entered the hull from undiscovered source/s and accumulated in the vessel's after compartments and engine-room. The total extent and weight of the accumulated floodwater are unknown; however, the after freeboard, reserve buoyancy and transverse stability were markedly reduced, such that the vessel's remaining righting ability was overcome in the prevailing rough weather conditions. The water accumulated in the engine-room immersed the batteries, making the radios unusable and precluding any emergency broadcasts. The DALEWOODPROVIDER was not operating as a small fishing vessel at the time of the capsizing, but was transporting deck cargoes of timber. The vessel's as-built and approved Trim and Stability Booklet was inaccurate and redundant. Revised stability data prepared and submitted shortly after the vessel's previous capsizing in 1989 were inapplicable because the previously proposed sponson additions to the hull sides, on which the data were based, had not been installed. The vessel's stability was not re-assessed after the removal of fishing gear and before her entering into operation as a cargo-carrying vessel. At the time of her loss, the vessel held a SIC 29 certificate for a commercial fishing vessel, issued by the Ship Safety Branch of the CCG on 13 July 1994 and valid until 22 June 1998. The skipper did not have any formal marine training. The skipper did not exercise good judgement in that he left two inexperienced persons at the con in heavy weather. Two of the three crew members drowned after the raft overturned and they could not reach shore safely. Autopsies revealed that the cause of death of both victims was by drowning. The overturned and stranded wreck of the vessel was considered non salvageable at the time of completion of this report.Findings Sea water entered the hull from undiscovered source/s and accumulated in the vessel's after compartments and engine-room. The total extent and weight of the accumulated floodwater are unknown; however, the after freeboard, reserve buoyancy and transverse stability were markedly reduced, such that the vessel's remaining righting ability was overcome in the prevailing rough weather conditions. The water accumulated in the engine-room immersed the batteries, making the radios unusable and precluding any emergency broadcasts. The DALEWOODPROVIDER was not operating as a small fishing vessel at the time of the capsizing, but was transporting deck cargoes of timber. The vessel's as-built and approved Trim and Stability Booklet was inaccurate and redundant. Revised stability data prepared and submitted shortly after the vessel's previous capsizing in 1989 were inapplicable because the previously proposed sponson additions to the hull sides, on which the data were based, had not been installed. The vessel's stability was not re-assessed after the removal of fishing gear and before her entering into operation as a cargo-carrying vessel. At the time of her loss, the vessel held a SIC 29 certificate for a commercial fishing vessel, issued by the Ship Safety Branch of the CCG on 13 July 1994 and valid until 22 June 1998. The skipper did not have any formal marine training. The skipper did not exercise good judgement in that he left two inexperienced persons at the con in heavy weather. Two of the three crew members drowned after the raft overturned and they could not reach shore safely. Autopsies revealed that the cause of death of both victims was by drowning. The overturned and stranded wreck of the vessel was considered non salvageable at the time of completion of this report. The DALEWOODPROVIDER capsized because sea water entered the hull from undiscovered source/s and accumulated in after underdeck compartments and the engine-room. The location, weight and free surface effect of the floodwater markedly reduced the after freeboard, reserve buoyancy and transverse stability, such that the vessel's righting ability was overcome in the prevailing rough weather conditions. Two persons drowned because they were thrown from their overturned liferaft and they could not swim to shore in the very rough seas.Causes and Contributing Factors The DALEWOODPROVIDER capsized because sea water entered the hull from undiscovered source/s and accumulated in after underdeck compartments and the engine-room. The location, weight and free surface effect of the floodwater markedly reduced the after freeboard, reserve buoyancy and transverse stability, such that the vessel's righting ability was overcome in the prevailing rough weather conditions. Two persons drowned because they were thrown from their overturned liferaft and they could not swim to shore in the very rough seas. Following the occurrence, TSB Marine Safety Advisory No. 10/96 was forwarded to TC-Marine concerning the potentially unsafe stability condition of two sister vessels of the DALEWOOD PROVIDER (the HARBOUR PROVIDER and PACIFIC HORIZON).Safety Action Taken Following the occurrence, TSB Marine Safety Advisory No. 10/96 was forwarded to TC-Marine concerning the potentially unsafe stability condition of two sister vessels of the DALEWOOD PROVIDER (the HARBOUR PROVIDER and PACIFIC HORIZON).